Forward tests

Strategies published up front and then tracked forward, wins and losses alike. The point is the gap between a backtest and what happens next.

Loading the track record…

Every strategy here was published before its forward window started, and nothing is removed when it goes badly. A backtest Sharpe in the last column that fails to show up in the forward columns is exactly the result this page exists to surface.

Forward track record

How to read this

Sharpe is annualised return per unit of risk over the forward window only — the backtest column is there for contrast, not as evidence.
B&H Sharpe is the same window for simply holding the asset. A strategy that does not clear this bar has earned nothing for the extra complexity.
Exposure is the share of days the strategy actually held a position; low exposure with a high Sharpe means very few observations, so treat it as noise.
Windows are short. Below roughly a year, none of these numbers separate skill from luck — that is the honest reading, and it is why the page shows days.

Snapshot in words — 23 September 2026, 00:10 UTC

29 strategies are on the public forward track, live since 29 June 2026, with 6.0 bps of cost charged to every trade. The longest forward window is 87 days; 0 of 29 currently beat buy-and-hold on Sharpe over their own forward window. Best forward Sharpe right now: etf_flow_trend (BTC) 3.69 against 3.71 for holding; cme_am_fade (BTC) 3.67 against 3.97 for holding; combo4 (BTC) 3.37 against 3.71 for holding. Every line was published before its forward window started and nothing is removed when it goes badly — a backtest Sharpe that fails to show up forward is exactly the result this page exists to surface.

This paragraph is a static snapshot, written into the page when the site was last built. The table above loads live and is recomputed daily. If the two disagree, the live numbers are the current ones.

Data is generated on a schedule from public market sources and shown as-is. Research product, not investment advice. 18+